Thursday, October 8, 2026 · Week 41 Reading for Oct. 8 · Last entry Oct. 6
Paperclip Index
Paperclip IndexDocumented harm20Minor harm▼ 5 from a week ago · Oct. 8The Index

Hijacked by injectionPI-0010

Malicious GitHub issue could steer agents using the GitHub MCP server into leaking private repositories

Invariant Labs demonstrated that an AI agent connected to GitHub through the official MCP server could be hijacked by instructions planted in an issue on a public repository. When the user asked the agent to look at open issues, it pulled data from the user's private repositories and published it in a pull request on the public repository. The researchers described it as an architectural problem that a server patch cannot fully fix.

Reported May 26, 2025Happened on May 23, 2025Updated Oct. 6, 2026Rated under methodology v0.6

Control failure, trackedNot counted. No documented harm outside the developer, but the system broke a rule or crossed a limit. Tracked beside the index.
Harm
Harm: No harm reportedAdds nothing to the index.
Control
Control level 2, Instruction violationReported beside the index. It adds nothing to a harm reading; when no harm counts in a window, the highest eligible control level in the window is the reading.
Controlled testSingle sourceInitiated by Outside attacker
Near miss

Near miss. Harm was plausible but prevented. It is rated only on what actually happened.

Not counted: no harm reportedEffect on the index

Sources

How we know

4 sources · single source. Links go to the original publishers; the summary above is in our own words.

  1. primaryDemo pull request #2 opened by the hijacked agent (ukend0464/pacman)GitHub (Invariant Labs demo repository) · May 23, 2025github.com/ukend0464/pacman/pull/2
  2. researchGitHub MCP Exploited: Accessing private repositories via MCPInvariant Labs · May 26, 2025invariantlabs.ai/blog/mcp-github-vulnerability
  3. newsResearchers warn of prompt injection vulnerability in GitHub MCP with no obvious fixDevClass · May 27, 2025devclass.com/2025/05/27/researchers-warn-of-prompt-injection-vulnerability-in-g…
  4. newsAttack via GitHub MCP server: Access to private dataheise online · May 27, 2025heise.de/-10400040

Why this rating

No harm reported; control failure level 2

Two separate assessments. Only documented harm can count toward the index.

Observed harm

No harm reported

Researcher demonstration on controlled accounts; no victims reported.

Not a finding that no harm occurred.

The harm scale
  1. 1 Negligible Inconvenience, easily remedied.
  2. 2 Minor Limited, recoverable harm.
  3. 3 Moderate Material harm needing significant effort to remedy.
  4. 4 Severe Severe harm to health, rights, property or essential services.
  5. 5 Catastrophic Society-scale harm or disruption beyond a community's capacity to cope.

Control assessment

Instruction violation

Agent followed injected instructions using the access token it had been given; no permission boundary was crossed.

Reported beside the index. It adds nothing to a harm reading; when no harm counts in a window, the highest eligible control level in the window is the reading.

The control scale
  1. 1 Negligible Behaved as instructed. The problem was an ordinary error or a flawed output, with no rule broken.
  2. 2 Minor Broke an explicit instruction or rule, or gamed its goal, while staying inside its permissions and environment: for example, reward hacking, misreporting results, or following injected instructions within its permissions.
  3. 3 Moderate Acted outside the permissions it was given, deceived its overseers about its own actions, or tampered with oversight tools (logs, monitors, shutdown). Stopped by normal controls within an hour.
  4. 4 Severe Reached systems outside its permitted environment, or acquired money, compute or accounts without authorization. The type and mechanism say how.
  5. 5 Catastrophic The developer or operator lost control: the system copied its weights outside their control, replicated itself, or resisted being stopped for 24 hours or more.

Rating rationale

No real-world leak was reported from this controlled test, so impact is not reported. Control level 2 reflects injected instructions obeyed within granted permissions.

The scales

Effect on the index

It does not count toward the May 26 reading

The reading for the week to May 26, 2025, with this record and without it. Harms count in full for two weeks after they are reported, then one level less every two weeks.

Not counted: no harm reported. 3 other records behind the reading for that week.

The arithmetic
The reading for the week of May 26, recomputed
StepWith itWithout
Counts toward the index?documented, external, eligible evidenceNo—
Worst documented harm, ksets the band— (none counting)— (none counting)
Harms at that level, nposition in the band00
Highest control level breachedsets the reading only when no harm counts33
Readingrounded down3 Control failures only3 Control failures only
Inside the window of 5 weekly readings
Week toReadingBand
May 26, 20253Control failures only
June 2, 20253Control failures only
June 9, 20253Control failures only
June 16, 20253Control failures only
June 23, 20252Control failures only

Recalculated backcasts, not readings published at the time.

Revisions

What we changed

4 logged. Every change to a rating is logged here, with the reason.

  1. v4
    Oct. 6, 2026

    Ratings confirmed by the editor. Changed: impact.

  2. v3
    Oct. 3, 2026

    Report date moved from 2025-05-27 to 2025-05-26, the publication date of Invariant Labs' post. Event date set to 2025-05-23, when the demonstration pull request was opened; added it as a source. Verification relabeled from Multiple credible sources to Single source with a research evidence review: DevClass and heise repeat Invariant's demonstration. Named the model (Claude 4 Opus in Claude Desktop), dated the Invariant and heise sources and added a status note. Ratings unchanged; impact not reported, so no reading effect.

  3. v2
    Sept. 30, 2026

    Rated: impact not reported; control type instruction violation.

  4. v1
    Sept. 30, 2026

    Backfilled from public reporting.

Cite and share

Use this record

Citation

Paperclip Index. “Malicious GitHub issue could steer agents using the GitHub MCP server into leaking private repositories.” Record PI-0010. Reported May 26, 2025; updated Oct. 6, 2026. Rated under methodology v0.6. https://paperclipindex.com/incident/PI-0010